Optical transmission systems in which optical pulses are transmitted over optical fibers, typically of low loss silica, are becoming of increasing importance. Generally these systems use wavelength division multiplexing (WDM) to increase the channel capacity of transmission thereby to reduce the unit cost per channel.
The lowest cost transmitters for WDM networks presently are directly modulated lasers, for example distributed feedback (DFB) lasers. However, the pulses produced by such lasers are generally characterized by chromatic dispersion in that the leading edge of the pulse typically includes frequency components that are changing from high to low whereas the trailing edge typically includes frequency components that are changing from low to high. Such pulses when transmitted over an optical fiber experience a phenomenon described as chirping, which leads to a reduction in the distance a train of closely spaced pulses can be transmitted without overlap between pulses. Such overlap impairs the fidelity of transmission and the recovery of the signal information at the receiver.
One technique that has been used to improve the quality of pulses from a directly modulated laser has been to pass the pulses through a narrow pass band filter to remove these unwanted frequencies at the leading and trailing edges of the pulses. However, in a WDM system using many channels of different wavelengths, this approach requires a separate filter for each channel and adds to the complexity and cost of the system.
Additionally, it has been known that following a non-return-to-zero (NRZ) electro-absorption modulator with a soliton pulse shaper attenuates the regions of highest transient chromatic dispersion generated by the NRZ modulator for soliton transmission systems. However, such a scheme has not previously been proposed for attenuating the region of highest chromatic dispersion in directly modulated lasers.